[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fXDhtxW31DOfwjALK54kQ5aLJL0rAqEIa6VbU2WPekqw":3},{"code":4,"msg":5,"data":6},200,"操作成功",{"id":7,"title":8,"content":9,"digest":10,"source":10,"coverPath":11,"thumbsCoverPath":12,"isTop":13,"isShow":14,"baseClick":13,"clickCount":15,"createTime":16,"typeId":17,"isNewest":18,"newsInfoTypeRespVo":19,"voiceUrl":22,"voiceSize":23,"taskId":24,"releaseTime":25,"titleEn":26,"contentEn":27,"voiceUrlEn":28,"taskIdEn":29,"voiceSizeEn":30},1352,"百度宣布文心X1.1深度思考模型上线 事实性比文心X1提升34.8%","\u003Cp class=\"ql-align-justify\">\u003Cstrong class=\"ql-lineHeight-1-75\" style=\"font-size: 18px; color: rgb(255, 153, 0);\">9月9日，在WAVE SUMMIT深度学习开发者大会2025期间，百度首席技术官、深度学习技术及应用国家工程研究中心主任王海峰正式发布了文心大模型X1.1深度思考模型，该模型在事实性、指令遵循、智能体等能力上均有显著提升。\u003C\u002Fstrong>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cspan style=\"font-size: 18px;\" class=\"ql-lineHeight-1-75\">当前，用户可以通过文心一言官网和文小言APP上体验并使用文心大模型X1.1。此外，文心大模型X1.1已正式登陆百度智能云千帆平台，面向企业客户以及开发者全面开放应用。\u003C\u002Fspan>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cspan style=\"font-size: 18px;\" class=\"ql-lineHeight-1-75\">据王海峰现场介绍，文心大模型X1是基于文心大模型4.5训练而来的深度思考模型，升级后的X1.1主要采用了迭代式混合强化学习训练框架，一方面通过混合强化学习，同时提升通用任务和智能体任务的效果；另一方面通过自蒸馏数据的迭代式生产及训练，不断提升模型整体效果。相比文心X1，X1.1的事实性提升34.8%，指令遵循提升12.5%，智能体提升9.6%。\u003C\u002Fspan>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cimg alt=\"undefined\" src=\"https:\u002F\u002Fimage.51xinwei.com\u002F2025\u002F09\u002F42f754a4c1d347d0a9bea81a7de1d3fb\u002Fbd398c3431804120b2541440e5bfee1b~tplv-tt-origin-web_gif.jpeg\" width=\"719\" height=\"undefined\" style=\"display: block; margin: auto;\">\u003Cp>\u003Cbr>\u003C\u002Fp>\u003Cp>\u003Cspan style=\"font-size: 18px;\" class=\"ql-lineHeight-1-75\">在多个权威基准评测中，文心X1.1整体表现超越DeepSeek R1-0528，在部分任务上展现出领先优势。同时，在与国际顶尖模型GPT-5和Gemini 2.5 Pro相比，效果持平。\u003C\u002Fspan>\u003C\u002Fp>\u003Cp>\u003Cbr>\u003C\u002Fp>\u003Cimg alt=\"undefined\" src=\"https:\u002F\u002Fimage.51xinwei.com\u002F2025\u002F09\u002F22572d80d7764950898a3f61159e20b7\u002Fa93d111d094f069ac53eb24fec12e2bb~tplv-tt-origin-web_gif.jpeg\" width=\"707\" height=\"undefined\" style=\"display: block; margin: auto;\">\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cspan style=\"font-size: 18px;\" class=\"ql-lineHeight-1-75\">文心大模型能够实现能力范围的拓展与运行效率的显著提升，这离不开飞桨文心二者联合优化所发挥的关键作用。在大会现场，百度重磅发布了飞桨核心框架3.2版本，该版本在大模型训练、硬件适配以及生态支持等多个关键领域实现了全方位升级。与此同时，百度还对大模型开发套件ERNIEKit和高效部署套件FastDeploy进行了同步升级。\u003C\u002Fspan>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cspan style=\"font-size: 18px;\" class=\"ql-lineHeight-1-75\">值得注意的是，百度方面的最新数据显示，飞桨文心生态开发者达到2333万，服务企业达到76万家。\u003C\u002Fspan>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cspan style=\"color: rgb(187, 187, 187);\">【新闻来源】环球网 \u003C\u002Fspan>\u003Ca href=\"https:\u002F\u002Fwww.toutiao.com\u002Farticle\u002F7547933508186964480\u002F?upstream_biz=doubao&source=m_redirect\" rel=\"noopener noreferrer\" target=\"_blank\" style=\"color: rgb(187, 187, 187);\">https:\u002F\u002Fwww.toutiao.com\u002Farticle\u002F7547933508186964480\u002F?upstream_biz=doubao&source=m_redirect\u003C\u002Fa>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cspan style=\"color: rgb(187, 187, 187);\">（本网转发此文章，旨在为读者提供更多的信息资讯，所涉内容不构成投资、消费建议。文章事实如有疑问，请与有关方核实，文章观点非本网观点，仅供读者参考。）\u003C\u002Fspan>\u003C\u002Fp>","","https:\u002F\u002Fimage.51xinwei.com\u002F2025\u002F09\u002F9426e3d6b6e342aea4b400851d35ee5a\u002FAI领域.jpg","https:\u002F\u002Fimage.51xinwei.com\u002F2025\u002F09\u002Fthumbs\u002F9426e3d6b6e342aea4b400851d35ee5a\u002FAI领域.jpg",0,1,45,"2025-09-12 15:06",2,false,{"id":17,"name":20,"enName":21},"芯位视野","Xinwei Vision","https:\u002F\u002Fxinwei-dev-test.oss-cn-shenzhen.aliyuncs.com\u002Fintelligent\u002Faudio%3A9901cae5-4192-49ba-ab08-862084c03645%3A0.wav?Expires=1757676330&OSSAccessKeyId=LTAI5tNvY2RkKjZw4LLWsrPK&Signature=2cxbe3daSbYEVpe7jb655ZPeRqU%3D",3995810,"9901cae5-4192-49ba-ab08-862084c03645","2025-09-12 15:03","Baidu announced the launch of the WENXIN X1.1 Deep Thinking Model, with factual accuracy improved by 34.8% compared to WENXIN X1.","\u003Cp class=\"ql-align-justify\">\u003Cstrong class=\"ql-lineHeight-1-75\" style=\"font-size: 18px; color: rgb(255, 153, 0);\">On September 9, during the WAVE SUMMIT Deep Learning Developer Conference 2025, Baidu's Chief Technology Officer and Director of the National Engineering Research Center for Deep Learning Technology and Applications, Wang Haifeng, officially launched the WENXIN Large Model X1.1 Deep Thinking Model. This model has seen significant improvements in factual accuracy, instruction following, and agent capabilities.\u003C\u002Fstrong>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cspan style=\"font-size: 18px;\" class=\"ql-lineHeight-1-75\">Currently, users can experience and use the WENXIN Large Model X1.1 through the WENXIN Yiyan official website and the WENXIAOYAN APP. In addition, the WENXIN Large Model X1.1 has officially been launched on the Baidu Intelligent Cloud Qianfan platform, providing comprehensive access for enterprise customers and developers.\u003C\u002Fspan>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cspan style=\"font-size: 18px;\" class=\"ql-lineHeight-1-75\">According to Wang Haifeng's on-site introduction, the WENXIN Large Model X1 is a deep thinking model based on the WENXIN Large Model 4.5. After upgrading, the X1.1 mainly adopts an iterative hybrid reinforcement learning training framework. On one hand, it improves the effectiveness of general tasks and agent tasks through hybrid reinforcement learning. On the other hand, it continuously enhances the overall model performance through iterative production and training of self-distilled data. Compared to WENXIN X1, the factual accuracy of X1.1 is improved by 34.8%, instruction following by 12.5%, and agent capabilities by 9.6%.\u003C\u002Fspan>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cimg alt=\"undefined\" src=\"https:\u002F\u002Fimage.51xinwei.com\u002F2025\u002F09\u002F42f754a4c1d347d0a9bea81a7de1d3fb\u002Fbd398c3431804120b2541440e5bfee1b~tplv-tt-origin-web_gif.jpeg\" width=\"719\" height=\"undefined\" style=\"display: block; margin: auto;\">\u003Cp>\u003Cbr>\u003C\u002Fp>\u003Cp>\u003Cspan style=\"font-size: 18px;\" class=\"ql-lineHeight-1-75\">In multiple authoritative benchmark evaluations, the WENXIN X1.1 performed better than DeepSeek R1-0528, showing a leading advantage in some tasks. At the same time, when compared with international top models such as GPT-5 and Gemini 2.5 Pro, its performance is comparable.\u003C\u002Fspan>\u003C\u002Fp>\u003Cp>\u003Cbr>\u003C\u002Fp>\u003Cimg alt=\"undefined\" src=\"https:\u002F\u002Fimage.51xinwei.com\u002F2025\u002F09\u002F22572d80d7764950898a3f61159e20b7\u002Fa93d111d094f069ac53eb24fec12e2bb~tplv-tt-origin-web_gif.jpeg\" width=\"707\" height=\"undefined\" style=\"display: block; margin: auto;\">\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cspan style=\"font-size: 18px;\" class=\"ql-lineHeight-1-75\">The ability expansion and significant improvement in operational efficiency of the WENXIN Large Model could not have been achieved without the key role played by the joint optimization of PaddlePaddle and WENXIN. At the conference site, Baidu officially released version 3.2 of the PaddlePaddle core framework, which achieved comprehensive upgrades in multiple key areas including large model training, hardware adaptation, and ecological support. At the same time, Baidu also synchronized the upgrade of the large model development kit ERNIEKit and the efficient deployment kit FastDeploy.\u003C\u002Fspan>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cspan style=\"font-size: 18px;\" class=\"ql-lineHeight-1-75\">Notably, the latest data from Baidu shows that the PaddlePaddle WENXIN ecosystem has reached 23.33 million developers, serving 760,000 enterprises.\u003C\u002Fspan>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cbr>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cspan style=\"color: rgb(187, 187, 187);\">[News Source] Global Times \u003C\u002Fspan>\u003Ca href=\"https:\u002F\u002Fwww.toutiao.com\u002Farticle\u002F7547933508186964480\u002F?upstream_biz=doubao&source=m_redirect\" rel=\"noopener noreferrer\" target=\"_blank\" style=\"color: rgb(187, 187, 187);\">https:\u002F\u002Fwww.toutiao.com\u002Farticle\u002F7547933508186964480\u002F?upstream_biz=doubao&source=m_redirect\u003C\u002Fa>\u003C\u002Fp>\u003Cp class=\"ql-align-justify\">\u003Cspan style=\"color: rgb(187, 187, 187);\">（This article is reprinted by this site to provide readers with more information. The content does not constitute investment or consumption advice. If there are any doubts about the facts of the article, please verify with the relevant parties. The views of the article are not the views of this site and are for reference only.）\u003C\u002Fspan>\u003C\u002Fp>","https:\u002F\u002Fxinwei-dev-test.oss-cn-shenzhen.aliyuncs.com\u002Fintelligent\u002Faudio%3A7e88a232-3f75-4059-8bf0-730e38811140%3A0.wav?Expires=1774838472&OSSAccessKeyId=LTAI5tNvY2RkKjZw4LLWsrPK&Signature=rBZT4B%2BJ4osxJFgC5VT3KGiT2jw%3D","7e88a232-3f75-4059-8bf0-730e38811140",5608812]